CR 68:183-199 (2016)  -  DOI: https://doi.org/10.3354/cr01396

ABSTRACT: In this paper we analyse the joint distribution of extreme temperature and heavy precipitation events in southern South America during 1961-2000, and the predominant atmospheric circulation associated with the occurrence of compound extreme events. We show that the probability of occurrence of intense precipitation (daily rainfall higher than the 75th percentile) significantly increases during or following a warm night (minimum temperature higher than the 90th percentile), but decreases during a cold night (minimum temperature lower than the 10th percentile) during the warm season. Heavy precipitation events are associated with the simultaneous occurrence of warm days (maximum temperature higher than the 90th percentile) or following such an event in eastern Argentina, but they rarely occur before. In contrast, cold days (maximum temperature lower than the 10th percentile) happen more often after an intense rain. Compound events are usually associated with 1 or 2 typical circulation patterns in each subregion. For example, warm days and heavy precipitation tends to occur more often when a trough over the Pacific Ocean and a cold front over the continent lead to warm and wet air advected to the east of the region of study. We also analysed the skill of 7 regional climate models from the CLARIS LPB project to simulate the statistical relationship between temperature and precipitation extremes in 1990-2000. Overall, models were able to simulate an increase in the probability of occurrence of extreme rainfall during warm nights and cold days, and an inhibition of precipitation during cold nights. However, models tend to fail to capture the spatial distribution of the compound extreme events in southeastern South America.

KEY WORDS: Compound extremes · Extreme temperature · Heavy precipitation · Atmospheric circulation · Regional climate models